In collaboration with 10 academic centers across the United States, we have recruited a large sample of families in which at least 2 siblings suffer from bipolar disorder or related mood disorders. This is the largest sample ever to participate in a genetic study of bipolar disorder. All research participants have undergone a diagnostic interview and provided a blood sample for DNA analysis. Genetic linkage studies have been performed using molecular markers evenly spaced across all chromosomes. These studies suggested several chromosomal regions may contain genes that contribute to bipolar disorder in these families. Ongoing work is aimed at identifying the actual genes involved. Using the latest genotyping chip technology and DNA pooling, we conducted the first genome-wide association study of bipolar disorder. The results implicated several genes, each of small effect, suggesting that bipolar disorder is a polygenic disease. Meta-analysis of independent case-control studies of bipolar disorder supported association with a cluster of genes on chromosome 3p21 and with 2 independent regions of the ANK3 gene implicated in previous studies. In past years, we have shown that models based on large numbers of markers can distinguish between cases and controls in independent datasets with high significance, but only modest predictive value. These studies further suggest that common alleles predisposing to bipolar disorder also predispose to major depression and schizophrenia, but not to neurological diseases such as Parkinson Disease. We have also shown that DGKH, a gene implicated in the first genome-wide association study of bipolar disorder that we published in 2007, but inconsistently replicated in subsequent studies, shows differential expression in post-mortem brain tissue taken from patients diagnosed with bipolar disorder. In our most recent meta-analysis of genome-wide association studies of bipolar disorder, we increased the sample size to nearly 18,000 through inclusion of worldwide samples. We found novel association signals near the genes TRANK1, LMAN2L, and PTGFR, along with further support for ANK3 and the 3p21 locus. We combined these findings and the other discoveries to date to model a future discovery trajectory for bipolar disorder. The model suggests that detection of novel loci by genome-wide association will remain a relatively slow process , growing at about 2 loci per 2000 case-control samples until leveling off around 56 loci at total sample sizes near 100,000. This model puts a lower limit on the number of BD loci that exist in the population and estimates that these loci together may account for about 4% of the variance in risk for bipolar disorder. In an effort to identify genetic variants that may have a much larger impact on individual risk, we are beginning large scale sequencing studies in selected populations, such as the Amish in Holmes County, Ohio. Large family sizes increase the opportunities for ascertaining extended relatives. Since most Amish marry others in their community, even individuals belonging to different nuclear families are often related, forming an extended kindred ideal for genetic studies. And since the Amish community was founded by relatively few original settlers, genetic diversity is reduced, which may mean smaller pools of risk alleles involved in common disorders such as bipolar disorder. Fortunately, individuals tend to stay in the area where they were born, facilitating longitudinal and offspring studies. There is a long history of BD studies among the Amish in Lancaster County, PA. Our work is complementary and not redundant since we use a different ascertainment strategy (distant relatives instead of extended families), draw cases from different Amish populations, and focus on sequencing as our main method of investigation. So far we have collected about 50 distantly related individuals from Amish communities in Ohio and Indiana. Genotype analysis with common genetic markers confirms that all individuals are related (at the 2nd to 3rd cousin level, on average), and that several chromosomal regions are shared in common among many cases. Exome sequencing is underway in the laboratory of our collaborator, David Goldstein. We are also searching for genetic markers that might help predict an individual's response to lithium, one of the most effective current treatments for bipolar disorder. To this end, we organized a large international collaboration, known as the Consortium on Lithium Genetics (ConLiGen), which aims to characterize lithium response in a large group of patients using reliable instruments, then perform a genome-wide association study. So far we have collected data and DNA from over 1000 cases and have genotyped these on genome-wide marker arrays. Analysis is underway.